There may be a billion neutron stars scattered across the Milky Way, and almost none of them have ever been seen. These ultra-dense remnants of exploded stars, each packing more mass than the Sun into a sphere roughly the size of a city, emit little or no light once they exhaust their rotational energy. They drift through the galaxy like dark ships without beacons, invisible to every optical and radio telescope ever pointed at the sky.
That could change as early as 2027. NASA’s Nancy Grace Roman Space Telescope, currently in final assembly and testing as of June 2026, is designed to detect these objects not by their light but by their gravity. A peer-reviewed study published in Astronomy & Astrophysics (A&A 707, A264) forecasts that Roman’s planned survey of the galactic center will catch the gravitational fingerprints of isolated neutron stars, potentially producing the first large-scale census of these stellar corpses and revealing what researchers at Lawrence Livermore National Laboratory have called a “hidden city” of dead stars buried in the Milky Way’s densest neighborhoods.
Weighing what you cannot see
The detection method is called astrometric microlensing, and it works like this: when a massive but invisible object, such as an isolated neutron star, drifts near the line of sight to a distant background star, its gravity warps spacetime enough to bend the background star’s light. That bending produces two effects. The background star temporarily brightens (photometric microlensing), and its apparent position on the sky shifts by a tiny but measurable amount (astrometric microlensing).
The positional shift is the prize. Unlike brightening alone, which reveals that something massive is there, the shift encodes the lens object’s mass directly. Astronomers can weigh something they have never seen and never will see in any conventional image.
This is not hypothetical. The Hubble Space Telescope has already pulled it off once. According to NASA, Hubble used astrometric microlensing to constrain the mass of an isolated dark compact object in the Milky Way, cataloged as OGLE-2011-BLG-0462. But that single detection required years of painstaking follow-up through Hubble’s narrow field of view. Roman is built to do the same work at an entirely different scale.
Why Roman changes the math
The engine behind that scale is Roman’s Galactic Bulge Time-Domain Survey, or GBTDS. According to NASA’s survey overview, the GBTDS will revisit its target fields roughly every 12 minutes during observing seasons, staring at the dense star fields toward the galactic center. Roman’s wide-field instrument captures a patch of sky about 100 times larger than Hubble’s in a single exposure, meaning each revisit records positional data for millions of stars simultaneously.
That combination of rapid cadence and enormous coverage is what makes routine neutron star detection plausible for the first time. Where Hubble needed years to track one event, Roman can monitor millions of potential microlensing alignments in parallel, catching events as they unfold and measuring the telltale positional wobble in near-real time.
The A&A 707 study put numbers behind this promise. The research team built simulations of the Milky Way’s stellar population, generated synthetic microlensing events, and ran them through Roman’s expected measurement precision and survey geometry. The result: quantitative forecasts of how many isolated neutron stars the GBTDS should detect and how precisely their masses can be measured.
As described in an institutional summary from Lawrence Livermore National Laboratory, whose researchers contributed to the study, a large sample of direct mass measurements would carry information about how each neutron star formed. Some arise from standard iron-core-collapse supernovae; others form through rarer channels such as electron-capture supernovae. A statistically meaningful census would let astrophysicists distinguish between competing models of stellar death and test how factors like metallicity and binary interactions shape the neutron star mass spectrum.
The uncertainties that remain
Projections are not guarantees. The biggest open question is how many neutron stars Roman will actually find. The simulations depend on assumed population models, and the true density of isolated neutron stars in the galactic bulge is not well constrained. Stellar crowding in those dense fields could introduce systematic errors in astrometric measurements when millions of stars overlap on the detector.
The Hubble precedent illustrates the risk. Two independent research teams analyzed the same Hubble data for OGLE-2011-BLG-0462 and reached different conclusions: one group classified the object as a low-mass black hole, the other argued it could be a neutron star. The discrepancy was driven partly by systematic errors in astrometric fitting in a crowded field. A further preprint examining the same event found that crowded-field systematics can shift mass estimates enough to change an object’s classification entirely.
Roman’s designers have built safeguards into the GBTDS: consistent cadence, stable pointing, and a uniform observing strategy that should reduce the kind of ad hoc systematics that complicated the Hubble case. The wide field also allows simultaneous measurement of large comparison samples, improving calibration of subtle detector and optical distortions. Still, no space telescope has attempted astrometric microlensing at this scale, and actual performance in the bulge’s most crowded regions will not be known until data starts flowing.
Interpretation carries its own uncertainties. Even if Roman delivers hundreds of neutron star masses, reading the resulting mass distribution requires theoretical models of supernova physics that are themselves under active revision. The relative contributions of core-collapse, electron-capture, and more exotic formation channels remain debated, especially at the low- and high-mass ends of the neutron star spectrum. Whether the data will clearly favor one formation pathway over another depends on the precision of individual measurements, the representativeness of the detected sample, and the robustness of the theoretical models used to decode the results.
What a surprise would mean
Perhaps the most exciting possibility is the one nobody can plan for. If Roman finds significantly more or fewer neutron stars than predicted, or if their masses cluster in unexpected ways, that would signal either gaps in current models of stellar evolution or unanticipated observational biases. Sorting out which explanation is correct will require cross-checks with other surveys and follow-up observations from ground-based observatories.
For now, the scientific case rests on solid ground. Astrometric microlensing has been demonstrated on a single object with Hubble. Roman’s survey design, documented in NASA’s mission planning, scales that proven technique to an unprecedented volume of stars. And the peer-reviewed simulation work translates those design parameters into concrete, testable expectations.
The claim that Roman will transform the search for isolated neutron stars is not speculation. It is a forecast built on a well-specified instrument, a demonstrated detection method, and detailed modeling of the galaxy. The exact yield, the precision of the mass measurements, and the clarity of the astrophysical conclusions all remain to be seen. But when Roman turns its wide eye toward the crowded heart of the Milky Way, it will be looking for a population of dead stars that, until now, has existed only in theory. Finding even a fraction of them would open a window into the life cycles of massive stars that no previous observatory could crack.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
